Using a composite of the Nationwide and Halifax indices, I forecast average house prices will fall a further 30% from December 2008 levels to £115,000.

One of the big uncertainties for the economy at present is the appropriate level for house prices. With the Halifax just announcing a 2.2% fall in house prices in December alone we apparently have no reasonable forecast of how low that prices can go. I therefore will try to forecast the following

1. The value of the lowest house prices.
2. When they will bottom out.
3. When we can expect to see an upturn.

I will revise my forecast as new data emerges.

A simple forecast is to look at the levels to which house prices fell in the last bust. The Nationwide, and the Halifax use different methodologies, which amount to similar things.

Using the Nationwide methodology

The Nationwide believes that the long-term trend is for a 2.9% real growth in house prices per annum. Using this methodology they produced the following graph for their November 2008 report.

At the in Q3 2008, average house prices were £165,000, or 109% of trend. The peaks were 132% of trend in Q3 2004 and 131% of trend in Q2 and Q3 2007.

This compares with the previous peak of 134% in Q2 1989. The low point relative to trend was not until nearly 7 years later in Q1 1996, when prices dipped below 70% of trend.

The actual low was in Q4 1992, when prices were 81% of trend.

Let us assume that house prices bottom out at 80% of trend, but do so after 10 to 12 quarters, instead of the 14 previously. That means house prices will reach around £127,000 at the end of 2009 to 2010. That is a 23% fall on end of November levels.

Using the Halifax methodology

The Halifax index recorded an average house price of £160,000 in December. This represented 4.44 times average earnings. The long-term average is around 4 times earnings, which would imply another 10% fall. However, the housing market bottomed out last time at 3.10 time average earnings. If this is reached in at the end of 2010 (with average earnings 6% higher), then the average house price will be 26% lower than December 2008 levels (27.6% lower than November) at £118,000.

An Opinion

The average of these two estimates s for house prices to fall a further 25% from the end of 2008 levels. This assumes we have the same pattern as the previous slump. This is a bold step to make. The previous slump in house prices was caused by a sharp rise in interest rates to combat 10% inflation, with interest rates reaching 15%. The differences are as follows.

1. The current slump is much sharper. Using the Halifax index, in the first 17 months of the previous slump average house prices as a ratio of average earnings slumped 15%. In the 17 months since the peak in July 2007 the decline has been 24%.
2.  In the last slump, much of the decline in employment was in that first 17 months of the decline in house prices. In this slump, much of the decline in employment will be in 2009. It will therefore prolong the period of decline as supply of houses onto the market continues to exceed supply. Furthermore the number of repossessions will increase the supply, unless banks (or the government) let the houses rather than sell.
3. The Nationwide calculation is based on an underlying trend of 2.9%. The magnitude of this trend could be exaggerated as

i)                   The last few years have seen increased competition in the market, leading to more products and the ease of switching (with discounts).

ii)                 It would include the sustained 10 year increase in prices. This boom has been unusually prolonged.

iii)               There would have been a one-off effect from the end of inflation. Although the real cost of a mortgage would reminded unchanged, a high and fluctuating rates of inflation (as in the 1970s and 1980s) did mean higher real costs in the early years and volatility of repayments as a percentage of income. Also, if people moved to a more expensive property before the mortgage was paid off, then would end up missing out on some of the devaluation in the repayments as a consequence of the inflation.

All this leads to the conclusion that basing the cost on some multiple of average income is better than having a notional long-term trend. This would make the peak of 5.84 times earnings 146% of trend. The Q3 2008 price is therefore 119% of trend, not 109%. Therefore a slump to 80% of trend would give a house price of £111,000, or 33% lower than in Q3.

The average of my two estimates therefore becomes a fall of 30% on end 2008 levels to around £115,000. This is conditional on the banks stabilizing, and unemployment peaking in around 12 months time, or at least most of the uncertainties in the economy clearing in that time.

At present there appears to be more negatives, that could make this lower than the last time, than there are positives.

Why the forecast may be too optimistic.

1. The current slide has occurred prior to the increase in unemployment that the recession will bring. A long period of uncertainty will leave both potential buyers and lenders cautious about entering the fray.
2. A 30% slide will leave a great number of people with negative equity. They will therefore be unable to move without having first made considerable payments
3. Inflation will be near zero, whereas in the early to mid-nineties it was in the 3-5% range. The real cost of housing and of the debt will not by reduced so quickly by inflation this time.
4. A feature of the past boom was the large number of people who bought-to-let. Many houses were bought not for the return from letting, but for the return from appreciating value. For instance, in Manchester a three bedroom semi-detached sold for around £150,000 at the peak, yet could be rented for £600 a month. This would only fund a mortgage of £70,000 to £90,000. There is potentially a lot of people who would like to realize their investments, so any upturn in prices, or even more buyers coming to the market may lead to a huge increase in the supply of houses for sale.
5. There is a prediction of 70,000 repossessions in 2009. This could be much higher if unemployment goes higher than forecast.
6. In the past 5 years, interest rates have been much lower than historically. Once the economy stabilizes, it will be necessary to raise interest rates to levels of the 1990s of 5% to 7.5% range. Will real rates higher, real house prices will be suppressed.
7. The lack of available credit may be prolonged, especially with banks being required to hold more capital. Banks will probably become much more conservative anyway in their lending for a few years. Also with a thinner market and flat house prices, the same risk policy may mean that banks would only lend at 80% of the valuation, whereas they would lend with less risk at 105% of the valuation is a booming market, with over 10% annual inflation and large volumes. The reason is simple. In a booming market, someone in financial difficulties can manage a quick sale for more than they paid for the property. With a flat, thin market, a quick sale can only be achieved at a considerable discount.

Why the forecast may be too pessimistic.

1. The upturn will arrive when unemployment has stop increasing, or at least when the job uncertainties have much reduced. In the 1980s, house prices started to rise in 1983, three years before the numerical peak in unemployment.
2. The very low rates on mortgages will enable people to make overpayments to quickly bring down the value of their debt. If they have repayment mortgages, the lower interest rates will mean that repayments will have a much larger element of capital repayment. Therefore negative equity may not be as prolonged as in the previous slump. For instance, before the UK crashed out of the ERM, interest rates were over 11%. Even after, in the late 1990s they were still 7% to 8%. Although with inflation at 2% to 4% real rates were lower, they are still considerably above the 4.5% currently quoted for a new mortgage, or around 3.5% for exiting mortgages.
3.  Government actions to help support people with difficulties meeting mortgage repayments could stop the sharpness of the dip. However, this may conversely prolong a slump. A deep slump may reinforce the impact of a general economic upturn in making housing much cheaper for new entrants to the housing market, or for those wishing to move to more expensive housing.